how does inclusive growth boost tax revenue mobilization?

HAL Id: halshs-01281914https://halshs.archives-ouvertes.fr/halshs-01281914

Preprint submitted on 3 Mar 2016

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

How Does Inclusive Growth Boost Tax RevenueMobilization?

Jean-Louis Combes, Rasmané Ouedraogo

To cite this version:Jean-Louis Combes, Rasmané Ouedraogo. How Does Inclusive Growth Boost Tax Revenue Mobiliza-tion?. 2016. �halshs-01281914�

https://halshs.archives-ouvertes.fr/halshs-01281914

https://hal.archives-ouvertes.fr

C E N T R E D ' E T U D E S E T D E R E C H E R C H E S S U R L E D E V E L O P P E M E N T I N T E R N A T I O N A L

SÉRIE ÉTUDES ET DOCUMENTS

How Does Inclusive Growth Boost Tax Revenue Mobilization?

Jean-Louis Combes

Rasmané Ouedraogo

Études et Documents n° 5

February 2016

To cite this document:

Combes J.-L., Ouedraogo R. (2016) “How Does Inclusive Growth Boost Tax Revenue Mobilization?”, Études et Documents, n° 5, CERDI. http://cerdi.org/production/show/id/1789/type_production_id/1

CERDI 65 BD. F. MITTERRAND 63000 CLERMONT FERRAND – FRANCE TEL. + 33 4 73 17 74 00 FAX + 33 4 73 17 74 28 www.cerdi.org

http://cerdi.org/production/show/id/1789/type_production_id/1

http://www.cerdi.org/

Études et Documents n° 5, CERDI, 2016

2

The authors Jean-Louis Combes Professor, CERDI – Clermont Université, Université d’Auvergne, UMR CNRS 6587, 63009 Clermont-Ferrand, France. E-mail: [email protected]

Rasmané Ouedraogo PhD Student in Economics, CERDI – Clermont Université, Université d’Auvergne, UMR CNRS 6587, 63009 Clermont-Ferrand, France. E-mail: [email protected] Corresponding author: Rasmané Ouedraogo

This work was supported by the LABEX IDGM+ (ANR-10-LABX-14-01) within the program “Investissements d’Avenir” operated by the French National Research Agency (ANR).

Études et Documents are available online at: http://www.cerdi.org/ed

Director of Publication: Vianney Dequiedt Editor: Catherine Araujo Bonjean Publisher: Mariannick Cornec ISSN: 2114 - 7957

Disclaimer:

Études et Documents is a working papers series. Working Papers are not refereed, they constitute research in progress. Responsibility for the contents and opinions expressed in the working papers rests solely with the authors. Comments and suggestions are welcome and should be addressed to the authors.

mailto:[email protected]

http://www.cerdi.org/ed


3

Abstract Despite high economic growth in the last decades, many developing countries remain into

poverty, income inequality and unemployment of young people, which have led some

countries to adopt inclusive growth strategies. In this paper, we show how inclusive growth

can boost tax revenue mobilization in developing countries. Effective public policies,

improving social cohesion are fundamental to meet citizens’ aspirations and therefore incite

them to comply with taxes. To this end, this study uses GMM techniques to deal with

endogeneity issue and spans 55 developing countries over the period of 1995-2010. We find

that inclusive growth has positive effect on tax revenue mobilization. This finding is robust to

various aspects of inclusive growth measurements and additional controls.

Keywords Inclusive growth; Taxes revenue; Inequality; Employment.

JEL codes

F43, D63, O11, H20


4

1. Introduction

Many countries are recognizing the urgency of enhancing tax resource mobilisation (TRM) in

a global climate of economic uncertainty and competition. The global crisis convinced policy-

makers in developing countries that the time had come to reduce their dependence on external

resources and their vulnerability to external financial shocks by addressing the potential of

TRM. TRM at a significant level is essential to solidify ownership over development strategy,

and to strengthen the bonds of accountability between governments and their citizens. In

effect, DRM provides “policy space” to developing countries which is often constrained under

the terms and conditions of external resource providers. According to Mascagni, Moore and

Mccluskey (2014), the governments in developing countries need additional financial

resources to address the huge development challenges they face. While great progress was

made in recent years towards achieving the Millennium Development Goals, a large

proportion of people in low-income countries still face poverty, inequality, unemployment of

young people, malnutrition, vulnerability to natural disasters and preventable diseases,

amongst others. Aid has certainly contributed to alleviating some of these issues, but it is

becoming increasingly clear that the development challenge requires increasing tax revenue.

Developing countries across the world always suffer from insufficient supply of internal

resources. Very low tax to gross domestic product (GDP) ratio is a common characteristic of

most of the developing countries. This situation remains contrasted with the high economic

growth rates that developing countries have experienced in recent decades. Economic growth

is expected to raise people’s income, create job and opportunities and therefore increase the

tax base. However, this expectation is sometimes hampered by the absence of straight

mathematical relationship between economic growth and the increase in the tax base. Indeed,

it must that the benefits of growth to be shared to a large part of the population in order to

meet citizens’ aspirations (availability of public goods, reduction of poverty and inequality)

and therefore incite them to comply with taxes. Previous literature focused on economic

structural characteristics and institutional quality as the main determinants of tax mobilisation

and failed to integrate tax payer expectations of government policies and their willingness to

pay taxes (see for a review Chelliah 1971, Stotsky and WoldeMariam 1997, Bird, Martinez-

Vazquez and Torgler 2008; Haque 2011; Crivelli and Gupta 2014). This papers aims to fill

the gap in previous literature by focusing on the effect of inclusive growth on the mobilization

of tax revenue. According to Alm et al. (1992), tax compliance increases with (perceptions of)

the availability of public goods and services. They suggest that governments can increase


5

compliance by providing goods that citizens prefer in a more efficient and accessible manner,

or by more effectively emphasizing that taxes are necessary for the receipt of government

services. Accordingly, the main concern of taxpayers is what they get directly in return for

their tax payments in the form of public services. Therefore, the implementation of an

effective and fair taxation system is essential in order to finance public policies given the fact

that there is strong relationship between citizens and the state. According to the concept of

resource bargain, the state negotiates resources to finance public policies and the citizens

expect in return for their tax payments in the form of public services (Fjeldstad and Semboja

2001; Moore 2004). Then, individuals may pay taxes because they value the goods provided

by the government, recognizing that their payments are necessary both to help finance the

goods and services and to get others to contribute (Fjeldstad and Semboja 2001). Regarding

this theory of fiscal exchange, tax compliance depends on the respect of this contractual

relationship and the policies of the government. If the fiscal exchange is enforced, that implies

that populations will benefit from growth though government policies. That is what expected

with inclusive growth policy. Inclusive growth ensures that the economic opportunities

created by growth are available to all-particularly the poor-to the maximum possible extent.

The growth process creates new economic opportunities that are unevenly distributed.

Inclusive growth means the participation of all in the tangible benefits of economic growth,

made possible mainly by job creation that can lead to inequality and poverty reduction.

Taxation plays a vital role in promoting citizenship and reciprocal relations between the

taxpayer and government. It is about encouraging people to make a contribution for which

they receive something in return. In Sierra Leone, for instance, public resistance to the

introduction of the Goods and Services Tax in 2009 (which is another name for value added

taxes (VAT)) forced the government and its key donors to link the new tax to improved

provision of services.

These discussions aforementioned reveal that inclusive growth can boost tax revenue

mobilization or tax compliance. This paper aims to study the relationship between inclusive

growth and tax revenue mobilization. While a very limited works studied the determinants of

inclusive growth (See Anand, Mishra and Peiris, 2013), there is no empirical study that

tackles the consequences or the benefits of inclusive growth for developing countries. This

paper is attempting to come to grips with the link between inclusive growth and tax revenue

mobilization and therefore it seeks to understand how to mobilize resource in order to finance

development. No doubts that effective public policies, improving social cohesion and the


6

nexus between revenues and public goods are fundamental to meet the aspirations of the

citizens and therefore encourage them to comply with taxes.

Using GMM estimators to deal with endogeneity issue, our results show that inclusive growth

has positive impact on tax revenue mobilization. This finding is robust to various controls and

alternative measurements of inclusive growth index and the dependent variable.

The rest of the paper is organized as follow. Section 2 presents the measure of the inclusive

growth index. Section 3 specifies the empirical strategy and the explanatory variables, while

Section 4 presents the results obtained from the estimates. In section 5, we analyze

transmission channels through which inclusive growth could affect tax revenue mobilization.

Section 6 concludes and draws up some policy recommendations.

2. Measurement of Inclusive growth

The notion of inclusive growth has captured a growing importance in the literature, reflecting

the desire to define and measure an index that includes most prominent indicators of long

lasting development policies. However, it is noteworthy that there is no consensus on how to

define inclusive growth and how to measure it. The definition of inclusive growth originates

from the notion of pro-poor growth. The term “inclusive growth” was coined by Kakwani and

Pernia 2000) and refers to the notion of participation in and benefitting from growth

processes. Therefore, it goes beyond the reduced concept of pro-poor growth that focused on

the level and distributions of income outcomes and takes into account non-income outcomes.

Since then, there is a battery of works which tried to capture the dimensions of an inclusive

growth process and conceptualize the patterns of inclusive growth (See for review Ali and

Son 2007, Bhalla 2007, Ianchovichina and Lundstrom 2009, Klasen 2010, McKinley 2010,

Rauniyar and Kanbur 2010, Blanke and others 2015). The most common of all these works is

that economic growth does not automatically translate into widely shared gains and inclusive

growth is mostly referred to poverty alleviation, inequality reduction and employment. In a

report on inclusive growth, the Organization for Economic Cooperation and Development

(OECD) (2012) identifies three problems that even the record levels of growth of the 1990s

and decade of 2000s failed to tackle: poverty, unemployment and inequality. Therefore,

inclusive growth has two dimensions: benefit-sharing (inequality and poverty) and

participation (employment). The benefit-sharing dimension looks into whether the process led


7

to a decrease in poverty and in income inequality. This segment of the analysis is aligned with

the concept of “relative pro-poor growth”, which differs from “absolute pro-poor growth” due

to the understanding that, for growth to be pro-poor, poor people’s income must grow faster

than that of wealthier people, resulting in a decrease in inequality (Grosse, Harttgen and

Klasen 2008). The participation dimension is the second significant dimension in

conceptualizing inclusive growth and in differentiating it from pro-poor growth. The

participation dimension looks into how the society is involved in the process, given that such

involvement is essential for promoting social coherence and for capacity-building, which are

crucial for the sustainability of an inclusive growth process. Analyzed in the economic sphere,

a participatory process can be thought of as characterized by generating employment for a

significant part of a country’s population. Such conceptualization of inclusive growth based

on three dimensions is in line with Ramos, Ali and Son (2007), Ramos, Ranieri and Lammens

(2013) who attempted to define and measure inclusive growth. Notwithstanding the

unconcluded concept of inclusive growth, some authors have tried to measure it. However,

likewise the absence of definition, the measurement of inclusive growth is still not settled.

Many researchers merely compare trends in economic growth, poverty line, income inequality

and social indicators (See Habito 2009) or construct a simple average of indicators (See

McKinley 2010, Ramos, Ranieri, and Lammens 2013). However, all of these attempts suffer

from serious shortcomings. First, an update to any attribute of a composite index causes the

index to be modified and the weights are highly arbitrary and sensitive to the choices of the

author. Second, a merely analysis of trends does not give us a sense of how much is growth

inclusiveness and then does not give us neither a clear picture nor very much of an idea on

how to proceed. In this paper, we follow closely Ali and Son (2007) who attempted to

construct inclusive growth index using a macro social mobility function, following the micro

literature on income distribution. This measure of inclusive growth is based on a utilitarian

social welfare function drawn from consumer choice literature, where inclusive growth

depends on two factors: (i) income growth; and (ii) income distribution or employment. More

specifically, Ali and Son (2007) propose an income equity index as:

𝑤 = 𝑦∗̅̅̅̅

�̅� (1)

Where 𝑤 is the equity index, �̅� is the average of income or employment rate and 𝑦∗̅̅ ̅

represents inclusive growth.


8

Equation (1) implies that income or opportunities are equitably (inequitably) distributed if 𝑤

is greater (less) than 1. For a completely equitable society, 𝑤 = 1. Rearranging, �̅�∗ = 𝑤 ∗ �̅�.

Inclusive growth requires increasing �̅�∗, which could be achieved by: (i) increasing �̅� , i.e

increasing average income through growth or the average level of opportunities; (ii)

increasing the equity index of income or the index of opportunities, 𝑤; or (iii) a combination

of (i) and (ii). Differentiating the above equation:

𝑑�̅�∗ = 𝑤 ∗ 𝑑�̅� + 𝑑𝑤 ∗ �̅� (2)

Where 𝑑�̅�∗ measures the change in the degree of inclusive growth. If 𝑑�̅�∗ > 0, growth is

more inclusive. Equation (2) allows us to decompose inclusive growth into income growth

and change in equity or opportunities in the society. The first term in the right side of equation

(2) is the contribution of increase in average income (keeping income distribution or average

opportunities constant) while the second term is the contribution of changes in the income

distribution or job opportunities (keeping the average income or employment rate unchanged).

Therefore, inclusive growth depends on the sign and the magnitude of the two terms. It is

noteworthy that the two terms can be as well as all positive (growth is inclusive) or negative

(growth is not inclusive) or a combination of different signs of the two terms. For this latter

case, the inclusiveness of growth will depend on which contribution outweighs the other. To

illustrate the evolution of inclusive growth, we rearrange equation (2) as:

𝑑�̅�∗

�̅�∗=

𝑑�̅�

�̅�+

𝑑𝑤

𝑤 (3)

This transformation decomposes inclusive growth into growth (�̅�) and percentage change in

equity or opportunities (𝑤).

3. Data and stylized facts

Our dataset includes 55 developing countries, only data availability restricted our sample (see

the appendix for a list of countries). As pointed out above, we focus on income inequality and

employment. We collect annual data for the period from 1995 to 2010. We define three-year

panel data: 1995-1997; 1998-2000; 2001-2003; 2004-2006; 2007-2010. This type of data is

preferable to pure time-series or cross-sectional data, as it marries possible inter-temporal


9

dynamics and important cross-country variation. Three-year averaging is, moreover, likely to

minimize non-systematic errors in the data. GDP per capita and tax revenue over GDP are

from IMF datasets (IFS and World Economic Outlook). Income inequality data refers to Gini

index extracted from Standardizing the World Income Inequality Database (SWIID). We take

the advantage of a recently-compiled cross-country dataset that distinguishes market (before

taxes and transfers) inequality from net (after taxes and transfers) inequality. In this study, we

consider the net income inequality index. Gini index is ranged between 0 and 100, with higher

values representing high unequal income distribution. To build the indicators of income

distribution used in the econometric estimations, we reverse the original indicator of Gini

Index by the following formula: Gi,t = (100 − Giniindex)/100. This transformation ranges G

between 0 and 1. On this basis, G increases with the improvement of the distribution of

income. As for employment, we extract data from International Labor Organization datasets.

We choose as indicator employment to population ratio which is the proportion of a country's

population that is employed. Ages 15 and older are generally considered the working-age

population. The series is harmonized to account for differences in national data and scope of

coverage, collection and tabulation methodologies as well as for other country-specific factors

such as military service requirements. The remaining variables include trade openness

(import plus exports in percentage of GDP), the share of agriculture value added in percentage

of GDP, natural resource rents in percentage of GDP, foreign aid per capita and public

expenditures on education over GDP extracted from World Development Indicators database,

the World Bank.

We generate inclusive growth index for each 3-year window. In figure 1, we present the

relationship between government revenue over GDP (excluded grants) and inclusive growth

defined with respect to income inequality and employment rate. As we can observe, there is

strong positive relationship between inclusive growth and government revenue. We will come

back with further econometric analysis.


10

Figure 1: Inclusive growth and government revenue, 1995-2010.

4. Econometric model

Given the strong inertia of government revenue, we use a dynamic specification. More

specifically, we estimate the following equation:

𝑅𝑒𝑣𝑖,𝑡 =∝ +𝛿𝑅𝑒𝑣𝑖,𝑡−1 + 𝛽𝐼𝑛𝑐𝑙𝑢𝑖,𝑡 + 𝜃𝑋′𝑖,𝑡 + 𝑣𝑖 + 𝜑𝑡 + 휀𝑖,𝑡 (4)

Where 𝑅𝑒𝑣𝑖,𝑡 represents government revenue for country 𝑖 in time 𝑡, 𝐼𝑛𝑐𝑙𝑢𝑖,𝑡 is inclusive

growth, 𝑋′𝑖,𝑡 stands for other explanatory variables. We include 𝑣𝑖 to control for unobserved

time-invariant country-level characteristics that are potentially correlated with government

revenue and 𝜑𝑡 to control for time-varying shocks that affect all developing countries.

Inclusive growth is suspected of endogeneity because of omitted variables bias and reverse

causality. Indeed, some variables could affect both revenue and inclusive growth. For

instance, given the high dependence of developing countries to external conditions, an

international crisis may reduce government revenue as well as its ability to achieve its

objectives, namely inclusive growth. Furthermore, endogeneity of revenue may be due to

reserve causality. For example, a country that experiences high inclusive growth may

mobilize taxes, while at the same time government uses mobilized resources to finance its

policies and distribute between citizens. To deal with endogeneity issues arising from

simultaneity bias and reverse causality, we use GMM estimators which are more suited for

dynamic panel data. Apart from endogeneity of government revenue, GMM estimators allow

us to correct for endogeneity of all right-hand side variables. There are two GMM estimators

commonly used: the difference-GMM estimator (Arellano and Bond, 1991) and the system-

10

20

30

40

50

-.05 0 .05 .1 .15

Inclusive growth-with employment as indicator

Government revenue over GDP Fitted values

10

20

30

40

50

-.1 -.05 0 .05 .1 .15

Inclusive growth-with inequality as indicator

Government revenue over GDP Fitted values


11

GMM estimator (Arellano and Bover, 1995; Blundell and Bond, 1998). For the difference-

GMM estimator, equation (4) is differenced in first order in order to remove country fixed

effects, and the first differentiated variables are instrumented by the lagged values of the

variables in level. As for the system-GMM estimator, both equations in levels and in first

differences are used in a system that allows the use of lagged differences and lagged levels of

the explanatory variables as instruments. Therefore, system-GMM estimator is an extension

of the difference-GMM estimator. In this paper, our preferred estimator is the system-GMM

because it has been highlighted that the lagged values of variables in level as it is done with

the difference-GMM estimator are sometimes poor instruments for variables in first

differences. To check the validity of the instruments, we use the Hansen test for

overidentifying restrictions and Arellano and Bond’s test that investigates that there is no

second-order serial correlation in the first-differenced residuals.

Explanatory variables

As for explanatory variables, we estimate our model using factors driving government

revenues. In selecting all of these variables, we follow closely the previous literature. More

specifically, beyond inclusive growth we include the following variables:

The level of development defined as the log of real GDP per capita “Log(GDPPC”. Indeed,

a higher level of development goes together with a higher capacity to pay and collect taxes, as

well as a higher relative demand for public goods and services (see Chelliah 1971, Bahl 1971,

Tanzi 1987, Gupta 2007, Pessino and Fenochietto 2010).

The share of agriculture value added in percentage of GDP “Agriculture”. It is expected to

be negatively associated with tax revenues because the sector is harder to tax (Stosky and

WoldeMariam 1997, Gupta 2007, Botlole 2010). Agriculture sector in developing sector is

characterized by a large number of small producers who sell their output in informal markets,

either to exchange for other goods, or for self-consumption. This coupled with poor or a non-

existent book-keeping record makes it notoriously difficult to tax.

Trade openness expressed as imports plus exports in percentage of GDP. Contrary to

agriculture sector, trade taxes are easier to collect, especially in developing countries.

Therefore, a high degree of openness is expected to generate a higher tax ratio (See Chelliah

1971, Botlhole 2010, Pessino and Fenochietto 2010, Keen and Pery 2013).


12

The total public expenditure on education as percent of GDP and represents the level of

education “Education”. Indeed, more educated people can understand better how and why it is

necessary to pay taxes. With a higher level of education compliance will be higher. Therefore,

it is expected a positive relationship between this variable and the level of tax effort (Pessino

and Fenochietto 2010, Keen and Pery 2013).

Total aid per capita “Aid”. Another concern is the potential tax displacement effect of aid.

It is often argued that an increase in aid inflows will lower the government’s incentives to

increase its tax effort, or even that tax revenues can be reduced due to policy reforms linked to

aid flows (McGillivray and Morrissey, 2001). We therefore include foreign aid per capita to

control for this potential tax displacement effect of aid.

Natural resource rents in percentage of GDP “natural_rents”. The effect of natural

resources on government revenue is ambiguous. Indeed, on the one hand a resource-rich

country can generate large taxable surplus (Gupta 2007), while on the other hand natural

resources might reduce the incentives of the governments for collecting taxes (Lim 1988 and

Martinez-Vazquez 2001).

Quality of governance: We include “polity2” index that measures the degree of democracy

in a country. It is widespread that quality of governance is a key factor driving tax

mobilization (Gupta 2007, Le, Moreno-Dodson and Bayraktar 2012).

5. Results

5.1 Baseline results

We present here baseline results obtained by using OLS with fixed effects, difference GMM

and system GMM. To check the importance of inclusive growth, we report on the one hand

the results for the effects of economic growth, income inequality and employment on

government revenue and on the other hand the effect of inclusive growth on government

revenue. This allows us to compare the results of the two estimates and therefore show how

inclusive growth more matters for tax mobilization. Table 1 reports results for the effects of

GDP growth, income inequality and employment on government revenue, while Table 2

reports results for inclusive growth. We observe that GDP growth is positively associated

with government revenue, while income inequality (except column 8) and employment

(except column 12) are not different from zero. However, Table 2 sheds light that the


13

coefficient associated with inclusive growth is strongly positive and significant in all columns

(except column 4). Furthermore, this coefficient is high than the one associated to GDP

growth in Table 1. Then inclusive growth more matters. In other words, an increase of 1

percent in inclusive growth results in a rise of government revenue ranging between 0.12 and

0.31 percent over GDP. When citizens benefit from growth, they are more likely to comply

with tax authorities and this results in an increase of revenue collected by the government.

This finding is consistent with the theory of fiscal exchange according to which compliance

increases with (perceptions of) the availability of public goods and services. Governments

may increase compliance by providing goods that citizens prefer in a more efficient and

accessible manner (Alm et al. 1992). By offering new opportunities to all and reducing

income inequality, government improves its accountability, increases tax bases and incites

citizens to comply with taxes. Furthermore, we observe that the coefficient associated with

inclusive growth measured with respect to income inequality is higher than when it is

measured with respect to employment. Therefore, it might be thought that inequality

reduction brings in more resources to the government than employment promotion.

As for the remaining variables, we observe that the level of education, aid and natural

resource rents are positively associated with government revenue, while the agriculture share

in percentage of GDP affects negatively government revenue. These findings are consistent

with previous works (Stosky and WoldeMariam 1997, Gupta 2007; Bird, Martinez-Vazquez

and Torgler 2008).


14

Table 1: Testing for effects of gdp growth, income inequality and employment

Lag dep. variable 0.106 0.0821 0.0989 0.0946 -0.138 0.327*** 0.350*** -0.107 0.701*** 0.726*** 0.656*** 0.717***

(0.136) (0.131) (0.137) (0.135) (0.0883) (0.0867) (0.0898) (0.0727) (0.0840) (0.0854) (0.0846) (0.0707)

GDP growth 0.0512* 0.0455 0.1011*** 0.0694** 0.0523* 0.0508**

(0.0277) (0.0278) (0.0296) (0.0272) (0.0273) (0.0224)

Gini_net -0.0726 -0.0555 -0.0122 -0.0995* 0.0184 0.0501

(0.0637) (0.0599) (0.0602) (0.0594) (0.0516) (0.0536)

Employment 0.0398 0.0251 0.508** -0.103 0.0763 0.197***

(0.137) (0.129) (0.201) (0.147) (0.0915) (0.0753)

Log(gdppc) 0.564 1.039 0.859 0.441 7.004*** 15.14*** 16.92*** 7.649*** -0.759 1.138 -1.843 -1.837**

(2.158) (2.248) (2.362) (2.285) (1.600) (3.187) (3.118) (1.368) (1.337) (1.353) (1.328) (0.934)

Agriculture -0.0264 -0.0163 -0.0152 -0.0316 0.0417 -0.107** -0.186*** 0.0539 -0.0493 0.0416 -0.142*** -0.178***

(0.0555) (0.0566) (0.0588) (0.0565) (0.0746) (0.0501) (0.0682) (0.0628) (0.0664) (0.0660) (0.0348) (0.0384)

Trade 0.0304 0.0362* 0.0347 0.0298 0.0214 0.0355 0.0444* 0.0132 0.0167 0.00826 0.0431*** 0.0301***

(0.0207) (0.0214) (0.0225) (0.0214) (0.0299) (0.0246) (0.0260) (0.0220) (0.0106) (0.0123) (0.0147) (0.00963)

Education 1.628*** 1.549*** 1.533*** 1.608*** 1.209*** 1.458*** 2.008*** 1.398*** 1.319*** 1.407*** 1.952*** 2.037***

(0.424) (0.407) (0.440) (0.438) (0.392) (0.362) (0.440) (0.288) (0.254) (0.256) (0.338) (0.238)

Aid 0.0263** 0.0273** 0.0235* 0.0290** 0.0220 -0.0120 -0.0287 0.00524 0.0414*** 0.0312** 0.0575*** 0.0505***

(0.0126) (0.0125) (0.0129) (0.0123) (0.0207) (0.0161) (0.0223) (0.0151) (0.0126) (0.0129) (0.0178) (0.0115)

Natural_rents 0.381*** 0.380*** 0.377*** 0.383*** 0.497*** 0.549*** 0.568*** 0.510*** 0.120*** 0.106*** 0.118*** 0.130***

(0.0583) (0.0596) (0.0587) (0.0593) (0.0583) (0.0532) (0.0719) (0.0364) (0.0269) (0.0274) (0.0287) (0.0278)

Polity2 -0.0891 -0.113 -0.0975 -0.0961 0.148 0.0576 -0.149 0.124 0.225** 0.165** 0.176* 0.204***

(0.104) (0.105) (0.115) (0.103) (0.125) (0.133) (0.148) (0.110) (0.107) (0.0711) (0.0988) (0.0632)

Constant 3.979 4.113 -0.297 6.383 5.153 -12.68 7.131 -2.220

(16.15) (17.40) (17.33) (16.23) (13.21) (14.69) (15.18) (9.310)

Observations 151 151 151 151 91 91 91 91 151 151 151 151

Countries 54 54 54 54 38 38 38 38 54 54 54 54

R-squared 0.697 0.694 0.691 0.699

AR(1): p-value 0,028 0,472 0,031 0,136 0,01 0,01 0,016 0,003

AR(2): p-value 0,518 0,772 0,429 0,259 0,799 0,609 0,944 0,685

Hansen test, p-value 0,37 0,125 0,291 0,55 0,603 0,648 0,914 0,72

Nb instruments 8 8 8 8 12 12 12 12

Note: Time effects are included in all the regressions. Robust standard errors in parentheses.

The dependent variable is government revenue in percentage of GDP.

***p<0.01, significant at 1% ; **p<0.05, significant at 5%; *p<0.10, significant at 10%

Fixed effects Difference-GMM System-GMM


15

5.2 Sensitivity analysis

In this section, we analyze some conditional factors that can affect the positive relationship

between inclusive growth and revenue mobilization. To this end, we consider some indicators

of policies that a country can pursue to achieve inclusive development. For instance, to

improve inclusive growth, developing countries can adopt policies such as building schools

and hospitals, training teachers and doctors, providing access to water, sanitation and

Table 2: Baseline results: effect of inclusive growth

Fixed effects Diff-GMM System-GMM Fixed effects Diff-GMM System-GMM

Lag dep. variable 0.122 -0.0249 0.689*** 0.121 -0.0234 0.732***

(0.137) (0.101) (0.0735) (0.142) (0.0927) (0.0766)

Inclusive_Growth 0.1235*** 0.3105*** 0.1424*** 0.1043 0.2738*** 0.1074*

(0.0391) (0.0574) (0.0489) (0.0684) (0.0757) (0.0602)

Log(gdppc) 2.022 12.97*** -1.321 1.089 11.51*** -0.856

(2.112) (3.068) (1.168) (2.408) (1.940) (0.953)

Agriculture -0.0333 -0.0355 -0.0834** -0.0252 0.0475 -0.0228

(0.0566) (0.0527) (0.0373) (0.0541) (0.0642) (0.0428)

Trade 0.0261 -0.0259 0.00543 0.0295 -0.0305 0.000797

(0.0187) (0.0212) (0.00965) (0.0217) (0.0345) (0.0106)

Education 1.562*** 1.062*** 1.197*** 1.605*** 1.686*** 1.280***

(0.410) (0.323) (0.204) (0.429) (0.340) (0.210)

Aid 0.0223* -0.00210 0.0404*** 0.0241* -0.00292 0.0317***

(0.0125) (0.0206) (0.0136) (0.0126) (0.0203) (0.0116)

Natural_rents 0.359*** 0.393*** 0.101*** 0.361*** 0.445*** 0.106***

(0.0539) (0.0664) (0.0235) (0.0567) (0.0677) (0.0292)

Polity2 -0.0653 0.145 0.0732 -0.0991 0.190 0.145*

(0.110) (0.122) (0.0814) (0.107) (0.133) (0.0746)

Constant -6.092 12.67 0.266 6.243

(16.01) (10.00) (18.18) (9.085)

Observations 151 91 151 151 91 151

Countries 54 38 54 54 38 54

R-squared 0.705 0.697

AR(1): p-value 0,321 0,014 0,197 0,009

AR(2): p-value 0,44 0,826 0,292 0,759

Hansen test, p-value 0,422 0,824 0,186 0,57

Nb instruments 8 12 8 12




Measurement with respect to inequality Measurement with respect to employment


16

transportation. Therefore, we interact inclusive growth index with the following variables:

social protection expenditures; primary school pupil-teacher ratio (the number of pupils

enrolled in primary school divided by the number of primary school teachers); the number of

hospital beds per 1,000 people; infant mortality rate; and paved roads in percentage of total

roads. Results are reported in Table 3.

In the last decade, social protection has emerged as a policy framework employed to address

poverty and vulnerability in developing countries. Therefore, social protection has a strong

focus on poverty reduction and on providing support to the poorest (de Haan 2000; Barrientos

and Hulme 2005). Such policy may reduce inequality and increases tax bases. Table 3 shows

that the interaction term with social protection is positive and significant (see column 1). This

finding implies that inclusive growth is likely to be most helpful to government revenue when

it is combined with social protection.

Developing countries are characterized by low levels of human development (education and

health) and their abilities to achieve the Millennium Development Goals (MDGs) require

enhanced actions. Education and health have to contend with the lack of teachers, nurses or

doctors, with an absence or too little material. To benefit from growth, no doubts that people

need access to education and health in order to become effective participants. Investments in

the two sectors are visible and citizens can easily assess government actions and then decide

their attitudes toward taxation. Results reported in Table 3 highlight that inclusive growth

does not contribute to resources mobilization in countries experiencing an increase in pupil-

teacher ratio and mortality rate. The interaction terms with the two variables are negative and

significant (see columns 2 and 4). However, when governments provide hospital beds, they

are likely to incite citizens to pay taxes (see column 3).


17

At last, we consider transportation issue. Developing countries suffer from a severe lack of

infrastructure. It is widely acknowledged that the infrastructure deficit is one of the key

factors preventing developing countries from realizing its full potential for economic growth

and competitiveness in global markets. Infrastructure can affect not only the quantum, but

also the spatial distribution of economic activity. For instance, investments in transportation

(1) (2) (3) (4) (5)

Lag dep. variable 0.796*** 0.907*** 0.762*** 0.684*** 0.740***

(0.0690) (0.0440) (0.0708) (0.0666) (0.0539)

Inclusive_Growth 0.0741 0.2050*** 0.0283 0.7101*** 0.1249***

(0.0658) (0.0586) (0.0547) (0.2622) (0.0300)

Inclusive*Social protection 0.0715***

(0.0165)

inclusive*Pupil-teacher ratio -0.0512***

(0.0156)

Inclusive*Bed 0.0305**

(0.0137)

Inclusive*Mortality -0.1487**

(0.0671)

Inclusive*Road 0.1471

(0.2597)

Log(gdppc) 3.470*** 1.980*** 2.802*** -0.946 -0.187

(0.817) (0.726) (0.737) (0.722) (0.890)

Agriculture -0.123*** -0.112*** -0.0768*** -0.0877*** -0.0223

(0.0244) (0.0219) (0.0285) (0.0275) (0.0297)

Trade 0.0362*** -0.00685 -0.00146 -0.0120 -0.000747

(0.00940) (0.00692) (0.00707) (0.00921) (0.00905)

Education 0.289 0.474** 1.391*** 1.109*** 1.347***

(0.227) (0.211) (0.180) (0.177) (0.158)

Aid 0.0241** 0.0395*** 0.0395*** 0.0616*** 0.0418***

(0.0117) (0.0112) (0.00829) (0.00963) (0.00773)

Natural_rents 0.0418 0.0380* 0.0803*** 0.0805*** 0.0670**

(0.0277) (0.0221) (0.0268) (0.0241) (0.0275)

Polity2 -0.0814 0.139** 0.200*** 0.0789 0.0352

(0.0886) (0.0661) (0.0713) (0.0572) (0.0631)

Constant 37.53*** 17.16*** 21.55*** 10.72* 0.749

(6.402) (6.406) (6.945) (6.499) (7.145)

Observations 120 134 116 151 103

Countries 43 49 49 54 43

AR(1): p-value 0.019 0.028 0.04 0.023 0.099

AR(2): p-value 0.896 0.729 0.306 0.76 0.567

Hansen test, p-value 0.725 0.647 0.752 0.862 0.889

Nb instruments 12 12 12 12 12




Table 3: Testing for sensitivity analysis-with inclusive growth measured with respect to income inequality


18

infrastructure may cause some economic activity to shift from the roads areas, as a result of

lower transportation. Such investments promote surely inclusive growth by allowing many

people to benefit from these infrastructures. Table 3 shows that the interaction term with the

percentage of paved roads is not different from zero.

5.3 Robustness checks

To check the validity of our findings we undertake a number of robustness exercises. Note

that we reported here results for inclusive growth measured with respect to income inequality.

Results for inclusive growth measured with respect to employment are similar and they are

reported in appendix.

(i) Testing for additional controls on baseline specification

We add further controls in the baseline specification in order to take into account other

variables likely to affect government revenue. These additional controls are terms of trade

“TOT”, remittances in percentage of GDP “remittances”, consumer inflation rate “inflation”,

public debt in percentage of GDP “debt”, conflict, and natural disaster “disaster”. Results are

reported in Table 4.

In first column we control for terms of trade shock1. Given the fact that developing countries

are the subject of external shocks, controlling for terms of trade shocks aims to test whether

our findings are robust to economic fluctuations. Several studies have found that changes in

the terms of trade-the price of exports relative to the price of imports-can account for half of

the output volatility in developing countries (Mendoza 1995, Kose 2002 and Broda 2003).

Even controlling for these fluctuations, we observe that inclusive growth still affects

positively government revenue. Furthermore, the coefficient associated with terms of trade

shock is not different from zero.

In column (2), we include migrants’ remittances. Beyond the private use of remittances, they

can play a role at macroeconomic level. Indeed, remittances increase recipient endowments

and therefore their capacity to pay taxes. Ebeke (2010) shows that remittances significantly

increase both the level and the stability of government tax revenue ratio in receiving

1 To measure the shock, we resort to the Hodrick-Prescott (H-P) filter, which generates a smooth trend and

stationary deviations. The deviation of terms of trade from its H-P-filtered trend will then capture the shock. We

set a smoothing parameter of 6.25.


19

countries. Our results reported in column (2) are consistent with this finding. Moreover, our

hypothesis about the effect of inclusive growth on government revenue remains in force.

To ensure that our finding is robust to the Olivera–Tanzi effect, we include inflation rate in

column (3). According to Olivera-Tanzi theory, an economic situation involving a period of

(1) (2) (3) (4) (5) (6) (7)

Lag dep. variable 0.814*** 0.790*** 0.829*** 0.853*** 0.787*** 0.797*** 0.806***

(0.0614) (0.0645) (0.0546) (0.0543) (0.0591) (0.0634) (0.0753)

Inclusive_Growth 0.1529*** 0.0942** 0.0943*** 0.0825*** 0.1297*** 0.1275** 0.1184*

(0.0483) (0.0422) (0.0319) (0.0289) (0.0368) (0.0524) (0.0649)

Log(gdppc) -1.735 1.968** 1.062* 1.594** 1.258** -1.034 -0.458

(1.072) (0.950) (0.640) (0.714) (0.636) (0.702) (0.609)

Agriculture -0.0717* -0.0642** -0.0238 -0.0248 -0.0321 -0.0279 -0.0125

(0.0421) (0.0307) (0.0216) (0.0254) (0.0300) (0.0393) (0.0406)

Trade 0.0172* 0.0292*** 0.00770 -0.00833 0.00118 -0.00836 -0.0101

(0.00937) (0.00826) (0.00698) (0.00952) (0.00912) (0.00969) (0.0102)

Education 1.203*** 1.323*** 0.944*** 0.633*** 0.877*** 0.632*** 0.708***

(0.215) (0.233) (0.114) (0.0972) (0.196) (0.189) (0.239)

Aid 0.0225* 0.0328** 0.0279** 0.0124 0.0301*** 0.0288** 0.0240*

(0.0128) (0.0137) (0.0140) (0.0143) (0.0117) (0.0129) (0.0133)

Natural_rents 0.0800*** 0.0934*** 0.0521** 0.0421* 0.0555** 0.0595** 0.0478

(0.0248) (0.0276) (0.0217) (0.0236) (0.0255) (0.0245) (0.0302)

Polity2 0.0519 0.154** 0.104* 0.172*** -0.0794 0.148** 0.238***

(0.0799) (0.0630) (0.0595) (0.0478) (0.0544) (0.0643) (0.0864)

ToT 0.0766 0.127*** 0.0483 0.00998 0.0742* 0.0216 0.00974

(0.0555) (0.0371) (0.0418) (0.0561) (0.0414) (0.0550) (0.0708)

Remittances 0.136*** 0.119** -0.0835 0.128** -0.0512 -0.0524

(0.0521) (0.0563) (0.0576) (0.0508) (0.0658) (0.0711)

Inflation 0.00938 0.0230 0.0457 0.0329 0.00122

(0.0324) (0.0349) (0.0385) (0.0327) (0.0423)

Conflict 0.624 -1.343 -0.415 0.236

(0.589) (0.895) (0.864) (1.218)

Disaster 0.0337 0.0414 -0.00139

(0.0286) (0.0318) (0.0285)

GDP_growth -0.0431 0.0191

(0.0284) (0.0375)

Gini_net 0.0427

(0.0399)

Constant 11.90 13.49 8.127 14.52*** 10.95** 10.51* 4.094

(9.459) (8.825) (4.958) (5.144) (4.867) (5.771) (4.930)

Observations 149 148 148 148 148 148 148

Countries 53 52 52 52 52 52 52

AR(1): p-value 0.007 0.008 0.008 0.007 0.008 0.006 0.009

AR(2): p-value 0.589 0.672 0.688 0.597 0.563 0.447 0.617

Hansen test, p-value 0.756 0.862 0.786 0.927 0.955 0.955 0.979

Nb instruments 12 12 13 13 13 14 14




Table 4: Testing for additional controls-with inclusive growth measured with respect to income inequality


20

high inflation in a country results in a decline in the volume of tax collection and a slow

deterioration of real tax proceeds being collected by the government of that country. Table 4

shows that we do not find Olivera-Tanzi effect, while inclusive growth remains important

driver of government revenue.

In columns (4) and (5), we include conflict and natural disaster. Indeed, both conflicts and

natural disasters destroy economies and dislocate societies and therefore make tax collection

difficult. Even controlling for the two types of difficult situations, inclusive growth still

affects positively government revenue.

In last columns (6) and (7), we control for GDP growth and income inequality index in order

to ensure that our results are not sensitive to these two variables. We find that not only the two

variables are not significantly different from zero but also inclusive growth remains strongly

significant and positive.

Overall, we conclude that our finding is robust to many additional controls. Then, inclusive

growth matters for resource mobilization in developing world through the increase in tax base

and the incentive to comply with taxes.

(ii) Testing for alternative main dependent variable

Up to now, we use total government revenue excluded grants as main dependent variable.

Government Revenues encompass social contributions (e.g. contributions for pensions, health

and social security), taxes other than social contributions (e.g. taxes on consumption, income,

wealth, property and capital), and other revenues. Here, we exclusively consider tax revenues

in percentage of GDP in order to take into account the real capacity to increase revenue

through taxation. Results are reported in Table 5. As we can observe, the coefficients

associated with inclusive growth are strongly significant and positive in all columns except

columns 2. However, these coefficients are smaller than those of table 2 and 4. This can be

due to the exclusion of other sources of revenues. As for the other independent variables, we

find an Oliveria-Tanzi effect (see column 4-8), while conflict and natural disasters are

negatively associated with tax revenue.


21

(1) (2) (3) (4) (5) (6) (7) (8)

Lag dep. variable 0.708*** 0.824*** 0.811*** 0.754*** 0.790*** 0.780*** 0.709*** 0.698***

(0.0582) (0.0468) (0.0478) (0.0285) (0.0247) (0.0181) (0.0258) (0.0262)

Inclusive_Growth 0.0853* 0.0712 0.1269*** 0.0763** 0.0593*** 0.0779*** 0.0986*** 0.1110***

(0.0465) (0.0466) (0.0304) (0.0316) (0.0187) (0.0277) (0.0238) (0.0254)

Log(gdppc) -0.101 1.414*** 1.383*** 0.846** 1.515*** 0.522** 0.121 0.186

(0.552) (0.468) (0.415) (0.377) (0.325) (0.219) (0.279) (0.307)

Agriculture -0.0584*** -0.0180 -0.0489*** -0.0165 0.0104 0.00860 -0.00253 -0.0127

(0.0193) (0.0251) (0.0187) (0.0153) (0.0124) (0.0123) (0.0168) (0.0219)

Trade 0.00796 0.00629 -0.0117 -0.00920 0.0193*** -0.00727 0.0111* -0.00857

(0.0123) (0.00957) (0.0116) (0.00810) (0.00716) (0.00678) (0.00587) (0.0104)

Education 0.000805 0.0552 0.0507 0.144* 0.161** 0.111 0.375*** 0.364***

(0.155) (0.128) (0.108) (0.0786) (0.0676) (0.125) (0.0840) (0.117)

Aid 0.0511*** 0.0485*** 0.0499*** 0.0308*** 0.0356*** 0.0279*** 0.0248*** 0.0265***

(0.0115) (0.0134) (0.0116) (0.00497) (0.00550) (0.00381) (0.00413) (0.00440)

Natural_rents 0.105*** 0.0873*** 0.0752*** 0.0652*** 0.0333** 0.0494*** 0.0545*** 0.0601***

(0.0193) (0.0181) (0.0183) (0.0170) (0.0155) (0.0158) (0.0110) (0.0141)

Polity2 0.198*** 0.196*** 0.192*** 0.0984** 0.0470 0.0390 0.0946*** 0.124***

(0.0619) (0.0596) (0.0549) (0.0476) (0.0388) (0.0433) (0.0298) (0.0294)

ToT -0.0568 -0.0422 -0.122*** -0.143*** -0.0708*** -0.0371* -0.0432**

(0.0405) (0.0332) (0.0276) (0.0228) (0.0200) (0.0199) (0.0196)

Remittances 0.0218 0.0716*** 0.0666*** 0.0725*** -0.00618 0.0138

(0.0338) (0.0172) (0.0258) (0.0261) (0.0309) (0.0314)

Inflation -0.0900*** -0.106*** -0.102*** -0.120*** -0.130***

(0.0108) (0.00957) (0.0125) (0.0127) (0.0154)

Conflict -1.292*** 0.230 -0.736** 0.535

(0.308) (0.276) (0.306) (0.379)

Disaster -0.0448*** -0.0448*** -0.0422***

(0.0107) (0.0128) (0.0130)

GDP_growth 0.0489*** 0.0523***

(0.0165) (0.0177)

Gini_net 0.0150

(0.0139)

Constant 3.758 -11.73*** -7.550** -5.220* -10.97*** -4.076** -1.446 -2.681

(4.321) (4.139) (3.241) (3.080) (2.612) (1.771) (2.206) (3.114)

Observations 147 145 144 144 144 144 144 144

Countries 55 54 53 53 53 53 53 53

AR(1): p-value 0.044 0.044 0.067 0.038 0.062 0.048 0.052 0.04

AR(2): p-value 0.217 0.217 0.179 0.178 0.176 0.196 0.186 0.192

Hansen test, p-value 0.525 0.525 0.735 0.358 0.338 0.378 0.897 0.961

Nb instruments 12 12 12 13 13 13 14 14


The dependent variable is tax revenue in percentage of GDP.


Table 5: Testing for alternative dependent variable-with inclusive growth measured with respect to income inequality


22

(iii) Testing for alternative measure of inclusive growth

We now change our measure of inclusive growth. So far, we employed Ali and Son (2007)’s

methodology to construct inclusive growth index. We test whether our results are robust to an

alternative measure of inclusive growth. We follow Ramos, Ranieri, and Lammens (2013) by

constructing a composite index combining the data on income inequality and employment

rate. We first compute the simple average of the Gini index and employment rate and then we

compare to GDP growth within each three-year window. Results reported in Table 6 do not

alter our previous findings. In other words, inclusive growth policies have positive effects on

tax revenue mobilization.

(1) (2) (3) (4) (5) (6) (7) (8)

Lag dep. variable 0.690*** 0.794*** 0.784*** 0.848*** 0.822*** 0.774*** 0.805*** 0.839***

(0.0741) (0.0715) (0.0684) (0.0516) (0.0564) (0.0603) (0.0593) (0.0889)

Inclusive_Growth 0.2164*** 0.1610** 0.1376** 0.1810*** 0.1894*** 0.2092*** 0.3107*** 0.3612***

(0.0670) (0.0685) (0.0677) (0.0381) (0.0463) (0.0498) (0.0887) (0.0957)

Log(gdppc) -1.248 1.766** 1.708** 1.101* 1.086* -0.683 -0.932 -0.763

(1.015) (0.896) (0.762) (0.584) (0.647) (0.577) (0.730) (0.715)

Agriculture -0.0855** -0.0622 -0.0587* -0.0272 -0.0101 -0.0147 -0.0178 6.73e-05

(0.0379) (0.0394) (0.0343) (0.0205) (0.0274) (0.0320) (0.0388) (0.0354)

Trade 0.0125 0.0120 0.0262*** 0.00234 -0.0113 -0.00467 -0.00702 -0.0101

(0.00844) (0.00887) (0.00716) (0.00465) (0.00865) (0.00818) (0.00949) (0.00946)

Education 1.194*** 1.245*** 1.414*** 1.019*** 0.819*** 0.891*** 0.873*** 0.936***

(0.181) (0.158) (0.191) (0.128) (0.151) (0.194) (0.167) (0.334)

Aid 0.0431*** 0.0215* 0.0302** 0.0267** 0.0179 0.0296** 0.0213 0.0143

(0.0121) (0.0128) (0.0128) (0.0110) (0.0128) (0.0116) (0.0143) (0.0123)

Natural_rents 0.0992*** 0.0825*** 0.0803*** 0.0302 0.0255 0.0376 0.0293 0.0357

(0.0288) (0.0291) (0.0279) (0.0215) (0.0239) (0.0259) (0.0282) (0.0393)

Polity2 0.111 0.0715 0.165*** 0.114*** 0.158*** 0.0667 0.226*** 0.263***

(0.0729) (0.0736) (0.0527) (0.0424) (0.0464) (0.0455) (0.0626) (0.100)

TOT 0.106** 0.153*** 0.0552 0.0599 0.0915** 0.0281 0.00632

(0.0512) (0.0424) (0.0352) (0.0481) (0.0427) (0.0507) (0.0596)

Remittances 0.137** 0.139*** 0.132** 0.161*** -0.0750 -0.0348

(0.0542) (0.0485) (0.0563) (0.0515) (0.0680) (0.0758)

Inflation 0.0116 0.0370 -0.0622* 0.00442 -0.0528

(0.0284) (0.0355) (0.0368) (0.0343) (0.0369)

Conflict 0.221 -1.116 -0.194 -0.430

(0.610) (0.964) (0.880) (0.994)

Disaster 0.0344 0.0267 0.0221

(0.0279) (0.0304) (0.0287)

GDP growth 0.0731** 0.0866***

(0.0366) (0.0306)

Gini net 0.0364

(0.0596)

Constant 10.99 12.43 11.23 8.467* 10.33** 6.746 9.307 5.576

(8.895) (7.963) (7.339) (4.384) (4.800) (4.707) (5.766) (6.947)

Observations 151 149 148 148 148 148 148 148

Countries 54 53 52 52 52 52 52 52

AR(1): p-value 0.012 0.006 0.006 0.007 0.011 0.010 0.011 0.011

AR(2): p-value 0.969 0.589 0.647 0.751 0.688 0.649 0.552 0.489


Nb instruments 12 12 12 13 13 13 14 14




Table 6: Testing for alternative measure of inclusive growth


23

7. Conclusion

In this paper, we estimate the effect of inclusive growth on tax revenue mobilization.

Effective public policies, improving social cohesion and the nexus between revenues and

public goods are fundamental to meet the aspirations of the citizens and therefore encourage

them to comply with taxes. This paper uses GMM techniques to deal with endogeneity issue

and covers 55 developing countries over the period of 1995-2010. We found that inclusive

growth has a positive effect on tax revenue mobilization. This finding is robust both for

inclusive growth measured with respect to income inequality and employment; various

additional controls, alternative measure of inclusive growth and the dependent variable.

Furthermore, this positive effect works in countries where inclusive growth is associated with

social protection and an increase in the number of hospital beds. However, infant mortality

rate dampens the gains of inclusive growth. To mobilize sufficient tax revenue for the post-

2015 International Development Agenda, our findings are important in terms of policy

recommendations. First, we call for the adoption of inclusive development strategies in

developing countries that meet citizen needs. Developing country incumbents should improve

public services through reforms and promote efficiency, transparency and accountability in

the use of public resources in order to ensure social inclusion of people. Employment and

combating income inequality should be promoted to allow everyone to participate in society

and the economy. Education and health are major sectors which require important

investments. Second, concerted efforts should be made to increase tax compliance and

therefore mobilize broad mix of resources.


24

References

Ali, I. and H. H. Son (2007). Measuring Inclusive Growth. Asian Development Review, Vol.

24, No. 1, pp.11–31.

Alm, J., G. H. McClelland and W. D. Schulze (1992). Why do people pay taxes? Journal of

Public Economics, 48, 21-38.

Anand, R., S. Mishra and S. J. Peiris (2013). Inclusive Growth: Measurement and

determinants. IMF Working Paper WP/13/135.

Arellano, M. and Bond, S. (1991). Some tests of specification for panel data: Monte Carlo

evidence and an application to employment equations. Review of Economic Studies, 58, pp.

277–297.

Arellano, M. and Bover, O. (1995). Another look at the instrumental variable estimation of

error component models. Journal of Econometrics 68, pp. 29–51.

Bahl, R. (1971). A Regression Approach to Tax Effort and Tax Ratio Analysis. International

Monetary Fund Staff Paper: 18, pp. 570-612.

Barrientos, A. and D. Hulme (2005). Chronic poverty and social protection: Introduction.

European Journal of Development Research, Vol. 17, No. 1, March, pp. 1–7.

Bhalla, S. (2007). Incluisve Growth? Focus on Employment. Social Scientist, Vol. 35, No.

7/8: 24–43.

Bird,R. M., Martinez-Vasquez,J. and Torgler,B. (2008). Tax Effort in Developing Countries

and High Income Countries: The Impact of Corruption, Voice and Accountability. Economic

analysis & policy. Vol. 38 (No.1 ).

Blanke J., Corrigan G., Drzeniek M., and Samans R. (2015). Benchmarking Inclusive Growth

and Development. Discussion Paper, World Economic Forum. Available at

http://www3.weforum.org/docs/WEF_Inclusive_Growth_Development.pdf.

Blundell, R. and Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel

data models. Journal of Econometrics, 87(1), pp. 115–143.

Botlhole, T.D. (2010). Tax effort and the determinants of tax ratio in Sub-Sahara Africa.

International Conference On Applied Economics – ICOAE 2010. University of Queensland.

Brisbane Australia

Broda, C. (2003). Terms of Trade and Exchange Rate Regimes in Developing Countries.

Journal of International Economics, vol. 63(1), 31-58

Chelliah, R. J. (1971). Trends in Taxation in Developing Countries. Staff Papers Vol 18, IMF

1971. (Washington: International Monetary Fund)

Crivelli, E. and S. Gupta. (2014). Resource Blessing, Revenue Curse? Domestic Revenue

Effort in Resource-Rich Countries. IMF Working Paper N°14/5. (Washington: International

Monetary Fund)


25

De Haan, A. (2000). Introduction: The role of social protection in poverty reduction. In T.

Conway, A. de Haan and A. Norton (eds.), Social Protection: New Directions of Donor

Agencies. Department for International Development, London.

Ebeke, H. C. (2010). Remittances, value added tax and tax revenue in developing countries.

CERDI, Etudes et Documents, E2010.30

Fjeldstad, O. H. (2001). Taxation, coercion and donors. Local government tax enforcement in

Tanzania. The Journal of Modern African Studies, 39, 289-306.

Gupta,S.A. (2007). Determinants of Tax Revenue Efforts in Developing Countries. IMF

Working Paper (WP/07/184. Washington D.C.).

Grosse, M., Harttgen, K. and Klasen, S. (2008). Measuring Pro-Poor Growth in Non-Income

Dimensions, World Development, Vol. 36, No. 6: 1021–1047.

Habito, C.F. (2009). Patterns of Inclusive Growth in Asia: Insights from an Enhanced

Growth-Poverty Elasticity Analysis. ADBI Working Paper Series, No. 145. Tokyo, Asian

Development Bank Institute.

Haque, A.A.K.M. (2011). Determinants of low tax efforts of developing countries.

Department of Business Law and Taxation, Monash University.

Ianchovichina, E. and Lundstrom, S. (2009). Inclusive Growth Analytics: Framework and

Application. Policy Research Working Paper, No. 4851. Washington, DC, World Bank.

Keen, M., and V., Perry. (2013). Understanding countries’tax effort. IMF Working Paper

N°/13/244 (Washington: International Monetary Fund).

Klasen, S. (2010). Measuring and Monitoring Inclusive Growth: Multiple Definitions, Open

Questions, and Some Constructive Proposals. ADB Sustainable Development Working Paper

Series, No. 12. Mandaluyong City, Philippines, Asian Development Bank.

Kose, M. Ayhan. (2002). Explaining Business Cycles in Small Open Economies: How Much

Do World Prices Matter? Journal of International Economics 56, no. 2: 299-327.

Le, T.M., B. Modeno-Dodson and N. Bayraktar (2012). Tax capacity and tax effort. Extended

Cross-Country Analysis from 1994-2009. Policy Research Working Paper 6252, the World

Bank.

Lim, D. (1988). Tax Effort and Expenditure Policy in Resource-Rich Countries, Economic

development policies in resource rich countries: 128-153.

Martinez-Vazquez, J. (2001). Mexico: An Evaluation of the Main Features of the Tax System.

International Studies Program Working Paper Series, International Studies Program, Andrew

Young School of Policy Studies, Georgia State University: 56 pages.

Mascagni, G., M., Moore and R., Mccluskey (2014). Tax revenue mobilisation in developing

Countries: issues and challenges. European Parliament. Available at

http://www.europarl.europa.eu/activities/committees/studies.do?language=EN

McGillivray, M., and O. Morrissey (2000). Aid illusion and public sector fiscal behaviour.

CREDIT Research Paper 00/9, University of Nottingham.

http://www.europarl.europa.eu/activities/committees/studies.do?language=EN


26

McKinley, T. (2010). Inclusive Growth Criteria and Indicators: An Inclusive Growth Index

for Diagnosis of Country Progress. ADB Sustainable Development Working Paper Series,

No. 14. Mandaluyong City, Philippines, Asian Development Bank.

Mendoza, E. G. (1995). The Terms of Trade, the Real Exchange Rate, and Economic

Fluctuations. International Economic Review 36, no. 1 (February): 101-37.

Moore, M. (2004). Revenues, state formation, and the quality of governance in developing

countries. International Political Science Review, 25, 297-319.

OECD (2012). Promoting inclusive growth: Challenges and policies. OECD, Paris

Olivera, J. H.G. (1967). Money, Prices and Fiscal Lags: A Note on the Dynamics of Inflation,

Banca Nazionale del Lavoro Quarterly Review, No. 82, pp. 258-268.

Pessino, C., and R. Fenochietto. (2010). Determining Countries’ Tax Effort. Hacienda

Pública Española/Revista de Economía Pública, Vol. 195, pp. 61–68.

Ramos, R.A., Ranieri, R. and Lammens, J.W. (2013). Mapping Inclusive Growth in

Developing Countries. IPC-IG Working Paper, No. 105. Brasília, International Policy Centre

for Inclusive Growth.

Rauniyar, G. and Kanbur, R. (2010). Inclusive Development: Two Papers on

Conceptualization, Application, and the ADB Perspective. Mandaluyong City, Philippines,

Asian Development Bank.

Stotsky, J. and WoldeMariam, A. (1997). Tax Effort in SubSaharan Africa. Working Paper

97/107: International Monetary Fund, Washington, DC.

Tanzi, V. (1978): Inflation, Real Tax Revenue, and the Case for Inflationary Finance: Theory

with an Application to Argentina, IMF Staff Papers, Vol. 25, No. 3, pp. 417-451.

Tanzi, V (1987). Quantitative Characteristics of the Tax Systems of Developing Countries. In

the theory of Taxation for Developing Countries, edited by David New and Nicholas Stern.

New York: Oxford University.


27

Appendix

(1) (2) (3) (4) (5) (6) (7) (8)

Lag dep. variable 0.732*** 0.779*** 0.799*** 0.829*** 0.870*** 0.825*** 0.735*** 0.805***

(0.0766) (0.0582) (0.0649) (0.0621) (0.0592) (0.0648) (0.0569) (0.0739)

Inclusive_Growth 0.1074* 0.1034 0.2031*** 0.1551*** 0.1992*** 0.1881*** 0.3458*** 0.3033***

(0.0602) (0.0711) (0.0734) (0.0380) (0.0464) (0.0493) (0.0852) (0.0874)

Log(gdppc) -0.856 -1.175 -1.346 1.372** 1.101** -0.934 0.0914 -1.117

(0.953) (0.906) (0.913) (0.615) (0.558) (0.704) (0.190) (0.724)

Agriculture -0.0228 -0.0175 -0.0533 -0.0268 -0.00268 -0.00800 -0.0400** 0.0103

(0.0428) (0.0391) (0.0371) (0.0229) (0.0252) (0.0365) (0.0187) (0.0338)

Trade 0.000797 -0.00703 0.0272*** 0.000317 0.0208** 0.0152** 0.0160* -0.0150

(0.0106) (0.0111) (0.0105) (0.00745) (0.00832) (0.00732) (0.00865) (0.0101)

Education 1.280*** 1.299*** 1.553*** 1.058*** 0.730*** 0.798*** 1.137*** 0.971***

(0.210) (0.201) (0.155) (0.131) (0.122) (0.194) (0.213) (0.276)

Aid 0.0317*** 0.0216* 0.0250** 0.0248** 0.0165 0.0219* 0.0126 0.0139

(0.0116) (0.0114) (0.0121) (0.0126) (0.0124) (0.0118) (0.0119) (0.0117)

Natural_rents 0.106*** 0.0834*** 0.0557** 0.0317 0.0126 0.0334 0.0440 0.0399

(0.0292) (0.0301) (0.0240) (0.0208) (0.0277) (0.0299) (0.0335) (0.0405)

Polity2 0.145* 0.103 0.166*** -0.0685 0.128** -0.0889 0.148** 0.182**

(0.0746) (0.0752) (0.0577) (0.0591) (0.0555) (0.0583) (0.0667) (0.0771)

ToT 0.0715 0.178*** 0.0536 0.0542 0.0498 -0.00204 -0.0260

(0.0629) (0.0603) (0.0411) (0.0456) (0.0482) (0.0509) (0.0615)

Remittances 0.140* 0.158** 0.139** 0.156** -0.0460 -0.0917

(0.0720) (0.0666) (0.0589) (0.0631) (0.0565) (0.0684)

Inflation 0.0361 0.0349 0.0309 0.0268 0.00972

(0.0373) (0.0361) (0.0399) (0.0350) (0.0317)

Conflict 0.555 -0.962 0.188 -0.00969

(0.636) (0.961) (0.883) (0.960)

Disaster 0.0354 0.0376 0.0200

(0.0282) (0.0318) (0.0377)

GDP_growth -0.0675 -0.0655

(0.0429) (0.0398)

Employment -0.00727

(0.0471)

Constant 6.243 7.894 7.331 10.62** 10.14** 9.113 0 10.09

(9.085) (7.762) (8.281) (5.020) (4.121) (5.835) (0) (6.609)

Observations 151 149 148 148 148 148 148 148

Countries 54 53 52 52 52 52 52 52

AR(1): p-value 0.009 0.013 0.008 0.007 0.006 0.008 0.006 0.009

AR(2): p-value 0.759 0.503 0.935 0.746 0.711 0.681 0.633 0.648


Nb instruments 12 12 12 13 13 13 14 14




Table A1: Testing for additional controls-with inclusive growth measured with respect to employment


28

(1) (2) (3) (4) (5) (6) (7) (8)

Lag dep. variable 0.708*** 0.782*** 0.745*** 0.752*** 0.776*** 0.814*** 0.771*** 0.785***

(0.0692) (0.0464) (0.0435) (0.0294) (0.0169) (0.0250) (0.0417) (0.0407)

Inclusive_Growth 0.1644*** 0.1571*** 0.1514*** 0.1812*** 0.1799*** 0.2110*** 0.1317*** 0.1394***

(0.0415) (0.0370) (0.0313) (0.0224) (0.0214) (0.0155) (0.0263) (0.0265)

Log(gdppc) -1.006 0.190 0.383 -0.218 0.152 -0.310 -0.415 -0.161

(0.665) (0.527) (0.338) (0.283) (0.362) (0.274) (0.311) (0.260)

Agriculture -0.0694*** -0.0571*** -0.0595*** -0.0164 0.00533 -0.0152 -0.0176 -0.0179

(0.0220) (0.0180) (0.0189) (0.0157) (0.0168) (0.0118) (0.0136) (0.0172)

Trade 0.0130 0.0260** 0.0285** 0.00435 0.0134* -0.00457 0.0191** 0.0216**

(0.0122) (0.0110) (0.0127) (0.00599) (0.00777) (0.00562) (0.00834) (0.00908)

Education 0.193 0.187 0.270** 0.354*** 0.482*** 0.376*** 0.320*** 0.353***

(0.165) (0.142) (0.108) (0.0977) (0.117) (0.0848) (0.124) (0.116)

Aid 0.0470*** 0.0476*** 0.0472*** 0.0224*** 0.0241*** 0.0193*** 0.0201*** 0.0127**

(0.0106) (0.0108) (0.00996) (0.00617) (0.00671) (0.00483) (0.00405) (0.00537)

Natural_rents 0.0904*** 0.0819*** 0.0887*** 0.0163 0.00890 -0.00535 0.0199 0.0128

(0.0233) (0.0180) (0.0185) (0.0172) (0.0106) (0.0193) (0.0176) (0.0243)

Polity2 0.169*** 0.228*** 0.212*** -0.0313 -0.00796 -0.0341 0.0650 0.0563

(0.0653) (0.0487) (0.0532) (0.0519) (0.0286) (0.0348) (0.0448) (0.0436)

ToT 0.0171 0.00662 0.0718*** 0.0810*** -0.00282 0.0335* 0.0616**

(0.0450) (0.0439) (0.0233) (0.0182) (0.0180) (0.0186) (0.0263)

Remittances -0.0439 -0.00395 0.0161 0.00620 0.0196 0.0547*

(0.0487) (0.0305) (0.0247) (0.0258) (0.0320) (0.0306)

Inflation -0.113*** -0.108*** -0.122*** -0.0991*** -0.102***

(0.00821) (0.00772) (0.00703) (0.00950) (0.00907)

Conflict -1.399*** -0.854*** -0.777*** -1.172***

(0.282) (0.291) (0.215) (0.370)

Disaster -0.0333*** -0.0215** 0.0158

(0.00957) (0.0108) (0.0158)

GDP_growth 0.0568 0.0615

(0.0598) (0.0614)

Employment -0.0214

(0.0214)

Constant 9.780* -2.867 -2.863 1.972 -1.687 1.700 3.773 2.740

(5.015) (4.412) (2.316) (2.234) (2.718) (1.983) (2.324) (2.630)

Observations 147 145 144 144 144 144 144 144

Countries 55 54 53 53 53 53 53 53

AR(1): p-value 0.041 0.031 0.049 0.035 0.034 0.057 0.053 0.037

AR(2): p-value 0.25 0.256 0.239 0.31 0.246 0.243 0.195 0.202


Nb instruments 12 12 12 13 13 13 14 14


The dependent variable is tax revenue in percentage of GDP.


Table A2: Testing for alternative dependent variable-with inclusive growth measured with respect to employment


29

(1) (2) (3) (4) (5)

Lag dep. variable 0.764*** 0.889*** 0.836*** 0.689*** 0.725***

(0.0526) (0.0432) (0.0615) (0.0575) (0.0614)

Inclusive_Growth 0.0893 0.1966*** 0.2415*** 0.05600 0.1064**

(0.0782) (0.0741) (0.0935) (0.0452) (0.0414)

Inclusive*Social protection 0.0743***

(0.0208)

inclusive*Pupil-teacher ratio -0.0737***

(0.0175)

Inclusive*Bed 0.0193**

(0.0095)

Inclusive*Mortality -0.0516***

(0.0149)

Inclusive*Road 0.1317

(0.3356)

Log(gdppc) 3.318*** 2.224*** -0.872 -1.214 -0.584

(0.999) (0.568) (0.745) (0.833) (0.727)

Agriculture -0.133*** -0.0534*** 0.0213 -0.106*** -0.0356

(0.0303) (0.0178) (0.0380) (0.0353) (0.0354)

Trade 0.0206*** 0.0141* 0.0235*** -0.000354 -0.000892

(0.00701) (0.00742) (0.00791) (0.00768) (0.0103)

Education 0.0408 0.823*** 1.450*** 0.878*** 1.291***

(0.244) (0.193) (0.212) (0.173) (0.158)

Aid 0.0334*** 0.0288*** 0.0117 0.0470*** 0.0391***

(0.0120) (0.00807) (0.00925) (0.00701) (0.00847)

Natural_rents 0.0756*** 0.0400* 0.0313 0.104*** 0.0659**

(0.0207) (0.0211) (0.0209) (0.0231) (0.0264)

Polity2 0.0633 0.0855* 0.115 0.136** 0.00642

(0.0530) (0.0505) (0.0706) (0.0688) (0.0667)

Constant 35.64*** 17.27*** 4.738 13.18* 5.113

(9.257) (4.568) (6.710) (7.796) (6.200)

Observations 120 134 116 151 103

Countries 43 49 49 54 43

AR(1): p-value 0.028 0.044 0.042 0.015 0.086

AR(2): p-value 0.906 0.664 0.792 0.932 0.589

Hansen test, p-value 0.588 0.47 0.704 0.664 0.753

Nb instruments 12 12 12 12 12




Table A3: Testing for sensitivity analysis-with inclusive growth measured with respest to employment


30

Table A4: Data sources

code Variable Data sources

Inclusive

growth

Inclusive growth measured with

respect to income inequality

and employment

Author's calculations. Income inequality

data are from SWIID (2013) and

employment from ILO datasets

Government

revenue

Government revenue in

percentage of GDP

IMF database

Tax revenue Tax revenue in percentage of

GDP

IMF database

Aid Aid per capita OECD-QWIDS datasets

Natural rents Total natural resources rents

(percentage of GDP)

World Development Indicators, the

World Bank (2014)

Remittances Remittances in percentage of

GDP


World Bank (2014)

GDPPC GDP per capita World Development Indicators, the

World Bank (2014)

Terms trade Terms of trade IMF database

Disaster Natural disaster total damage International disaster database of the

Centre for Research on the Epidemiology

of Disasters (CRED)

Conflict Conflict. Dummy variable that

takes one if the country is in

conflict and zero otherwise

Uppsala Conflict Database of the

International Peace Research Institute

(PRIO)

Trade Imports plus exports in

percentage of GDP


World Bank (2014)

Inflation Consumer inflation rate World Development Indicators, the

World Bank (2014)

Agriculture Agriculture value added in

percentage of GDP


World Bank (2014)

Education Public spending on education in

percentage of GDP


World Bank (2014)

Polity2 Polity2 Index Polity4 Project (Integrated Network for

Societal Conflict Research (INSCR)

2013)

Social

protection

Social protection expenditures

in percentage of GDP


World Bank (2014)

Pupil-

teacher ratio

Number of pupils enrolled in

primary school divided by the

number of primary school

teachers


World Bank (2014)

Bed Number of beds per 1,000

people


World Bank (2014)

Mortality Infant Mortality rate World Development Indicators, the

World Bank (2014)

Road Paved roads in percentage of

total roads


World Bank (2014)


31

Table A5: Samples

Countries

Albania Lesotho

Argentina Madagascar

Armenia Mexico

Burundi Mali

Burkina Faso Mongolia

Bangladesh Mauritius

Belarus Malaysia

Bolivia Namibia

Brazil Niger

Bhutan Nicaragua

China Nepal

Cameroon Pakistan

Colombia Panama

Costa Rica Peru

Dominican Republic Philippines

Ecuador Paraguay

Egypt Senegal

Guatemala El Salvador

Indonesia Thailand

India Tajikistan

Iran Tunisia

Jordan Turkey

Kazakhstan Tanzania

Kenya Uganda

Kyrgyz Republic Ukraine

Cambodia Venezuela

Sri Lanka Yemen, Republic of